Characterization, Pollution Sources, and Health Risk of Ionic and Elemental Constituents in PM2.5 of Wuhan, Central China

Atmospheric PM2.5 samples from Wuhan, China were collected during a winter period of February and a summer period of August in 2018. The average PM2.5 mass concentration in winter reached 112 μg/m3—about two-fold higher than that found in summer. Eight ionic species constituted 1/3 of PM2.5, whereas more than 85% represented secondary ionic aerosols (NO3−, SO42− and NH4). Higher ratios of NO3−/SO42− (0.95–2.62) occurred in winter and lower ratios (0.11–0.42) occurred in summer showing the different contribution for mobile and stationary sources. Seventeen elemental species constituted about 10% of PM2.5, with over 95% Na, Mg, Al, Ca, Fe, K and Zn. Higher K-concentration occurred in winter indicating greater contribution from biomass and firework-burning. Carcinogenic risks by Cr, As, Cd, Ni and Pb in PM2.5 indicated that about 6.94 children and 46.5 adults among per million may risk getting cancer via inhalation during surrounding winter atmospheric sampling, while about 5.41 children and 36.6 adults have the same risk during summer. Enrichment factors (EFs) and elemental ratios showed that these hazardous elements were mainly from anthropogenic sources like coal and oil combustion, gasoline and diesel vehicles.


Introduction
Crowned by domestic economists as "China's economic and geographic center", Wuhan City is the most populated megacity in Central China, with an area of 8494 km 2 and a population of approximately 10.2 million. Situated at the junction of the Yangtze River and Hanjiang River, Wuhan covers a critical geographic location about 950 km north of Hong Kong, 700 km west of Shanghai, 1000 km south of Beijing and 980 km east of Chengdu [1,2]. Wuhan's gross domestic product (GDP) was more than 1.1 trillion yuan in 2015 [1,2]. With growing energy consumption economic development and rapid urbanization, Wuhan has been suffering from serious atmospheric pollution [1][2][3][4]. PM 2.5 -defined as particulate matter with aerodynamic diameters below 2.5 µm-usually is an intuitive evaluation indicator of air quality [3,4]. According to the report from AQI (https://www.aqistudy.cn/historydata/), the annual PM 2.5 in Wuhan was 88.0 µg/m 3

Determination of 8 Species of Ionic Contents and 17 Species of Elemental Content in PM2.5
The concentration of eight water-soluble ionic species (Cl − , NO3 − , SO4 2− , Na + , NH4 + , K + , Mg 2+ , Ca 2+ ) in PM2.5 samples were measured based on ion chromatography method reported from Ministry of the Environment, Japan (https://www.env.go.jp/air/osen/pm/ca/manual.html). Each sample was analyzed three times by liquid chromatograph (ICS1600, Dionex Aquion, Thermo Fisher Scientific CO, Waltham, MA, USA). The concentration of 17 elemental species (Na, Mg, Al, K, Ca, V, Cr, Mn, Fe, Ni, Cu, Zn, As, Se, Cd, Pb and Sb) in PM2.5 were based on acid decomposition/ICP-MS method reported by the Ministry of the Environment, Japan (https://www.env.go.jp/air/osen/pm/ca/manual.html). They were then analyzed by an inductively coupled plasma mass spectrometer (Agilent 7700, Agilent Technologies, Inc., Santa Clara, CA, USA). The detailed extraction and analysis methods followed those reported in our previous articles [17]. All statistical analyses of related data were completed using Microsoft Excel 365 software [17,18].

Backward Air Mass Trajectory Analysis
Backward air mass trajectory analysis is often used to get more information about the effects of transport patterns of air masses, and can be used to evaluate the impacts to air quality. In this study, each backward air mass trajectory was calculated by HYSPLIT model on the National Oceanic and

Determination of 8 Species of Ionic Contents and 17 Species of Elemental Content in PM 2.5
The concentration of eight water-soluble ionic species (Cl − , NO 3 − , SO 4 2− , Na + , NH 4 + , K + , Mg 2+ , Ca 2+ ) in PM 2.5 samples were measured based on ion chromatography method reported from Ministry of the Environment, Japan (https://www.env.go.jp/air/osen/pm/ca/manual.html). Each sample was analyzed three times by liquid chromatograph (ICS1600, Dionex Aquion, Thermo Fisher Scientific CO, Waltham, MA, USA). The concentration of 17 elemental species (Na, Mg, Al, K, Ca, V, Cr, Mn, Fe, Ni, Cu, Zn, As, Se, Cd, Pb and Sb) in PM 2.5 were based on acid decomposition/ICP-MS method reported by the Ministry of the Environment, Japan (https://www.env.go.jp/air/osen/pm/ca/manual.html). They were then analyzed by an inductively coupled plasma mass spectrometer (Agilent 7700, Agilent Technologies, Inc., Santa Clara, CA, USA). The detailed extraction and analysis methods followed those reported in our previous articles [17]. All statistical analyses of related data were completed using Microsoft Excel 365 software [17,18].

Backward Air Mass Trajectory Analysis
Backward air mass trajectory analysis is often used to get more information about the effects of transport patterns of air masses, and can be used to evaluate the impacts to air quality. In this study, each backward air mass trajectory was calculated by HYSPLIT model on the National Oceanic Mass concentrations of all investigated chemical constituents, major water-soluble ions and PM 2.5 in each sampling case during winter and summer sampling periods in 2018 are shown in Figure 2. The concentration of each elemental is shown in Table S1 and ionic content is shown in Table [5]. Simultaneously, the ratio of PM 2.5 /PM 10 was always selected to analyze the particle origin, possible health effects and formation process [2]. The values of PM 2.5 /PM 10 shown in Table  S3 were 0.71 (0.45-0.96) in winter and 0.61 (0.50-0.71) in summer, which was slightly higher than the mean ratios (0.56) of 190 cities in China [2,20]. This indicates that a relatively serious degree of fine particle pollution occurred in Wuhan. Pearson product-moment correlation coefficient analysis shown in Table S4 also shows that the mass PM 2.5 concentrations showed a strong correlation with atmospheric pollutants (winter, SO 2 , r = 0.74, CO, r = 0.69, p < 0.01; summer, O 3 , r = 0.55, p < 0.01). Figure 2 shows that ionic contents formed a large proportion of the PM 2.5 . The average proportion was 37.5% (25.8%-63.5%) in winter and 26.3% (16.6%-42.9%) in summer, indicating that water-soluble ions form a very important content of PM 2.5 . This seems very consistent with that reported from the previous articles: Zhang et al. (2011) found that the average ratio of water-soluble ions to PM 2.5 in Xi'an China from March 2014 to March 2015 was about 38.9% [21]. Lai et al. (2016) also found this ratio in PM 2.5 of Guangzhou from March 2012 to February 2013 was as high as 44.8% [22]. The Pearson product-moment correlation coefficient shown in Table S4 shows that mass PM 2.5 concentrations had a strong correlation with ionic contents (winter, r = 0.832, p < 0.01; summer, r = 0.657, p < 0.01). For seventeen determined elemental contents, the average was 6.32 µg/m 3 in winter and 3.50 µg/m 3 in summer. It appears that elemental contents were in a relative light concentration (2.65%−14.2%) of PM 2.5 during both sampling periods. The elemental ratio of PM 2.5 was consistent with about 10% referred from several relevant articles [1,13,23]. The following section presents more important information about ions and elements.

Chemical Characteristics of Major Water-Soluble Ionic Species
Average of 8 species water-soluble ionic species concentrations in PM 2.5 shown in Figure 3 was 41.0 µg/m 3 during the winter sampling period and 10.4 µg/m 3 during the summer sampling period which indicated that these in winter was about 4-fold higher than that in summer. Ionic contents were in a widely range from 21.6 µg/m 3 (20 February) to 70.2 µg/m 3 (10 February) during the winter sampling period and from 5.57 µg/m 3 (20 August) to 24.4 µg/m 3 (23 August) in summer. During the winter sampling period, Figure 3a shows that SO 4 2− , NO 3 − and NH 4 + called secondary ionic aerosols were the absolutely dominant component with proportion of 23.03%, 38.45% and 25.01%, the others were in order of Cl − (5.70%) > K + (4.54%) > Ca 2+ (2.81%) > Na + (0.24%) > Mg 2+ (0.22%). During the summer sampling period, ionic species shown in Figure 3b were in order of SO 4 2− (55.3%) > The proportion of secondary ionic aerosols to total ions was 86.5% in winter and 89.9% in summer. These values appear very consistent with previous articles. All indicate that secondary ionic aerosols also play a very important role in PM 2.5 in Wuhan during winter and summer of 2018.  reported these ratios was over 85% in Wuhan [1]; Yao (2002) showed these ratios were about 85% in Beijing and 80% in Shanghai [24]. The Pearson product-moment correlation coefficient shown in Table  S4 show that the mass ionic contents had strong correlation with secondary ionic aerosols (winter, NH 4 + , r = 0.97 (p < 0.01), NO 3 − , r = 0.94(p < 0.05); SO 4 2− , r = 0.56(p < 0.01); summer, NH 4 + , r = 0.99(p < 0.01), NO 3 − , r = 0.58(p < 0.01); SO 4 2− , r = 0.99(p < 0.05)). The K + ion in PM 2.5 was usually considered as a diagnostic tracer for biomass burning and vegetation [3,17,24]. It also had another important source from fireworks containing KNO 3 , KClO 3 and KClO 4 as oxidizers [25,26]. This may explain why ratios of K + ions in winter were as high as 4.54%, and the mean K + concentration was 1.86 µg/m 3 -about 7.75-fold higher than in summer, especially from 15 February to 18 February. The Na + ion-always considered common from sea-salt-was in a lower concentration in both winter and summer [13,24,[27][28][29].
ions in winter were as high as 4.54%, and the mean K concentration was 1.86 μg/m -about 7.75-fold higher than in summer, especially from 15 February to 18 February. The Na + ion-always considered common from sea-salt-was in a lower concentration in both winter and summer [13,24,[27][28][29]. Ionic balance can be used as an indicator for revealing the acidity balance of species ions in ambient particles by the ionic equivalent ratio of anions to cations. If the value is close to 1.00, it suggests that the ionic content is in a neutralized stage; in acidic stage with a value over 1.00 and in alkaline with a value below 1.00 [13,[30][31][32][33]. Figure 3b shows that ratios of anions-eq to cation-eq were in range of 0.67 to 0.80 in winter except a 1.44 occurrence on 19 February, and also about 0.65 to 0.82 in summer. The deviation of 1.00 may be caused by the lack measurement of CO3 2− , PO4 3− , F − , NO2 − and other anions. Another factor may because of the great K + enrichment. The slope of linear fitting lines was up to 1.27 (R 2 = 0.79) in winter and 1.20 (R 2 = 0.97) in summer indicating the complicated ionic contents and serious air pollutions [4].   Ionic balance can be used as an indicator for revealing the acidity balance of species ions in ambient particles by the ionic equivalent ratio of anions to cations. If the value is close to 1.00, it suggests that the ionic content is in a neutralized stage; in acidic stage with a value over 1.00 and in alkaline with a value below 1.00 [13,[30][31][32][33]. Figure 3b shows that ratios of anions-eq to cation-eq were in range of 0.67 to 0.80 in winter except a 1.44 occurrence on 19 February, and also about 0.65 to 0.82 in  in Wuhan during 2014 was 0.32 in summer and 0.83 in winter [4] while Wu (2019) found that this ratio was 0.48 in summer and 1.39 in winter of Wuhan [3]. This seasonal variation was also found in Shanghai [24]. Higher ratio and higher NO 3 − concentration in this article could be explained by the fast-developing transportation and the relevant emissions in Wuhan. Lower ratio and lower NO 3 − concentration could be explained by generated via the conversion of gaseous precursors. SO 2 and NOx in atmosphere could be got directly affects from kinds of meteorological factors and emission sources [3]. Sulfur oxidation ratio (SOR) and nitrogen oxidation ratio (NOR) were selected as indicator to know more process about the conversion of gaseous precursors. Sulfur oxidation ratio (SOR = [SOR]/([SOR] + [SO 2 ])) is defined as the molar ratio of sulfur in SO 4 2− to the total sulfur in SO 4 2− and SO 2, NOR ) is the molar ratio of nitrogen in NO 3 − to the total nitrogen in NO 3 − and NO 2 [36,37] . It has already been confirmed that the photochemical oxidation of SO 2 would occur when the SOR over 0.10 [3,38]. Figure 3d shows that SOR in both winter and summer was over 0.20, which indicated gas-particle conversion in the atmosphere occurred in each period, and NOR was about 0.21(0.10-0.38) in winter and 0.03(0.02−0.10) in summer. Higher values of NOR and SOR in winter indicated that more NO 2 and SO 2 may have oxidized to the nitrate and sulfate in atmosphere, especially in winter. Meanwhile, higher SOR and lower NOR in summer could be explained by the ammonium sulfate formation was more favored in summer with high temperature and sulfate competed with nitrate for ammonium [3]. Figure 3e shows that the slopes of regression fit between (NO 3 − + SO 4 2− )-equivalent and NH 4 + -equivalent was 1.29 in winter and 1.17 in summer, which indicates that the main forms of these species were chiefly NH 4 NO 3 and (NH 4 ) 2 SO 4 in both winter and summer [3,4,38]. With several ionic analyses, it was found that the PM 2.5 in Wuhan suffered the serious effects from coal combustion and vehicle emissions. It is very important to get more information about the inorganic contents and their health risk assessment.

Variation of Concentrations, Possible Sources of Elemental Species
Here, seventeen kinds of elements were selected to analyze composition changes, possible sources and possible health risk assessment in both sampling periods. Table 1 and Table S2 show that the average mass concentration of all seventeen elemental species was 6324.1 (1734-10,545) ng/m 3 during the winter sampling period, while it was 3460 (2080-5070) ng/m 3 during the summer sampling period. Th elements Na, Mg, Al, Ca, Fe, K, Zn were in a relative higher average concentration of over 100 ng/m 3 in both winter and spring sampling periods. The mass proportion of these elements to total elements in each sampling case were over 95%. The mass concentrations were in the order of K >200 ng/m 3 > Al > 1000 ng/m 3 > Fe> Na > Mg> Ca > Zn in winter and Al > Fe> K > 500 ng/m 3 > Na > Zn > Mg > Ca in summer. Generally, Mg, Al, Ca, Fe and K elements are considered as crustal elements mainly from natural soil and construction dust [1,39]. It is noteworthy that the K element in PM 2.5 from 13 February to 18 February was over 2000 ng/m 3 . The K element even exceeded 6000 ng/m 3 on 16 February, which could be explained by the burning of not only biomass but also large amounts of fireworks during 2018 Chinese New Year period [25]. The average sum of other 10 elemental species defined as trace elements was 208 ng/m 3  Heavy metal elements in PM 2.5 showed the serious health risks and they are considered too main from kinds of anthropogenic sources [1,2,13,40], including the represent industries, biomass burning, coal combustion and the vehicle emissions [13][14][15]. Enrichment factors (EFs) was widely selected as an indicator to distinguish the possible sources of elements and evaluate the anthropogenic effects on the relative elements [1,13,41]. Here, the Al element was selected as the crustal typical to calculate EFs of each element. The detailed calculation method may be found in the previous article [1,13]. The EFs was calculated as following: where C is the element concentration. The EFs of each element in PM 2.5 are shown in Table S5. According to the elemental contents, four pollution cases were selected as the typical to analyze the possible sources: The 16 February case with greatest concentrations; the 20 February case with the lightest concentrations during the winter sampling period; the 25 August case with greatest concentrations and the 27 August case with the lightest concentrations during the summer sampling period. When EFs are greater than 1.00, the element is mainly generated by kinds of anthropogenic sources, such as coal combustion and vehicle emission. When EFs fall between 1.00 to 10.00, the element is mainly from both anthropogenic sources and natural sources (soil source). When EFs are close to 1.00, it means that the elements are from a crustal origin.  Figure 4 clearly shows that EFs of Sb, Pb, Cd, Se and Zn were over 100 in all 4 pollution cases, which indicated they were generated by anthropogenic sources. The EFs of Cu and As were also over 10 indicating their anthropogenic sources. The EFs of Cr and Ni were over 10 on 20 February, 29 August and the 18 August case. The EFs of K on 16 February and 20 February were greater than that occurred on 25 August and 27 August showed that K element in winter received more anthropogenic sources (such as biomass burning and fireworks burning). The EFs of Mn on 16 February and 20 February cases were lower than that in other cases which indicated that Mn element in winter may be got more natural sources (such as soil and dust). The EFs of V were between 1.00 to 10.0 and the values in relative lighter sampling were a little greater which indicated more anthropogenic sources in these sampling cases.
Atmosphere 2020, 11, x FOR PEER REVIEW 9 of 16 and natural sources (soil source). When EFs are close to 1.00, it means that the elements are from a crustal origin. Figure 4 clearly shows that EFs of Sb, Pb, Cd, Se and Zn were over 100 in all 4 pollution cases, which indicated they were generated by anthropogenic sources. The EFs of Cu and As were also over 10 indicating their anthropogenic sources. The EFs of Cr and Ni were over 10 on 20 February, 29 August and the 18 August case. The EFs of K on 16 February and 20 February were greater than that occurred on 25 August and 27 August showed that K element in winter received more anthropogenic sources (such as biomass burning and fireworks burning). The EFs of Mn on 16 February and 20 February cases were lower than that in other cases which indicated that Mn element in winter may be got more natural sources (such as soil and dust). The EFs of V were between 1.00 to 10.0 and the values in relative lighter sampling were a little greater which indicated more anthropogenic sources in these sampling cases. The elemental ratio method is often selected as a tool to evaluate the profiles sources, origin of air masses [42,43]. The values of several ratios (As/V, V/Ni, Zn/Pb and Zn/ Cd) are shown in Table  S2. Generally, Zn, Cu, Pb and Cd elements are considered as traffic traces, while Cu and Zn are main from oil combustion and Cd may be from the transportation activities such as the brake and tire wear [44]. V, Cr, Mn, Ni and Cd elements may be got the effects form the stack emissions [15], Pb element may be emitted from the smelting and coal combustion [15,41]. The Se element is considered as an important coal combustion tracer [42,43]. In 16 February case, ratio of As/V was 2.24 indicating the great contribution of gasoline and diesel combustion sources [43,45]; ratio of Zn/Pb was 1.85 and Zn/ Cd was 153.7 indicated the main sources may be form gasoline vehicles; In 20 February case, As/V(0.61) indicated the main sources may be form gasoline vehicles, V/Ni(0.39) indicated the important source of oil burning, Zn/Pb(11.73) and Zn/ Cd (577.4) indicated the diesel vehicles and metal scrap incineration sources [43,45]. In the 18 August case, As/V(5.15) indicated the sources of coal combustion, V/Ni (0.49) indicated the oil burning source, Zn/Pb(6.75) indicated the diesel vehicles effects and Zn/Cd (252.3) indicated oil burning sources. Refs. [43,45]. In the 29 August case, As/V(1.55) indicated the sources from gasoline vehicles and gasoline combustion, V/Ni (0.09) indicated the main gasoline and diesel vehicles source, Zn/Pb (5.24) indicated the effects from diesel vehicles, Zn/ Cd (137.2) indicated oil burning sources [43,45]. With these elemental ratios, elemental concentrations and NO3/SO4 2− in these four pollution cases, it could be considered that diesel and gasoline combustion and vehicles may be the most important source on 16 February. The PM2.5 on 20 February were likely kinds of anthropogenic source, while vehicle emission may be the main source due to the greater Cu and Zn mass concentrations. Coal, oil and diesel combustion may be the main  The elemental ratio method is often selected as a tool to evaluate the profiles sources, origin of air masses [42,43]. The values of several ratios (As/V, V/Ni, Zn/Pb and Zn/ Cd) are shown in Table  S2. Generally, Zn, Cu, Pb and Cd elements are considered as traffic traces, while Cu and Zn are main from oil combustion and Cd may be from the transportation activities such as the brake and tire wear [44]. V, Cr, Mn, Ni and Cd elements may be got the effects form the stack emissions [15], Pb element may be emitted from the smelting and coal combustion [15,41]. The Se element is considered as an important coal combustion tracer [42,43]. In 16 February case, ratio of As/V was 2.24 indicating the great contribution of gasoline and diesel combustion sources [43,45]; ratio of Zn/Pb was 1.85 and Zn/ Cd was 153.7 indicated the main sources may be form gasoline vehicles; In 20 February case, As/V(0.61) indicated the main sources may be form gasoline vehicles, V/Ni(0.39) indicated the important source of oil burning, Zn/Pb(11.73) and Zn/ Cd (577.4) indicated the diesel vehicles and metal scrap incineration sources [43,45]. In the 18 August case, As/V(5.15) indicated the sources of coal combustion, V/Ni (0.49) indicated the oil burning source, Zn/Pb(6.75) indicated the diesel vehicles effects and Zn/Cd (252.3) indicated oil burning sources. Refs. [43,45]. In the 29 August case, As/V(1.55) indicated the sources from gasoline vehicles and gasoline combustion, V/Ni (0.09) indicated the main gasoline and diesel vehicles source, Zn/Pb (5.24) indicated the effects from diesel vehicles, Zn/ Cd (137.2) indicated oil burning sources [43,45]. With these elemental ratios, elemental concentrations and NO 3 /SO 4 2− in these four pollution cases, it could be considered that diesel and gasoline combustion and vehicles may be the most important source on 16 February. The PM 2.5 on 20 February were likely kinds of anthropogenic source, while vehicle emission may be the main source due to the greater Cu and Zn mass concentrations. Coal, oil and diesel combustion may be the main source for the 18 August case while gasoline and diesel vehicle may be main source on 29 August. All of them were suffered the effects from anthropogenic sources. Backward trajectory of air masses in these 4 pollution cases were calculated by NOAA HYSPLIT model to analyze the possible effects from hyperactive air mass and shown in Figure 5.
Atmosphere 2020, 11, x FOR PEER REVIEW 10 of 16 source for the 18 August case while gasoline and diesel vehicle may be main source on 29 August. All of them were suffered the effects from anthropogenic sources. Backward trajectory of air masses in these 4 pollution cases were calculated by NOAA HYSPLIT model to analyze the possible effects from hyperactive air mass and shown in Figure 5.  Figure 5a shows that air masses on 16 February may be mainly from northeast and southeast. Figure 5b shows that air masses on 20 February were main from the middle-lower Yangtze River District. The other vital impact factors may be continuous raining. Figure 5c shows that air masses on 18 August were transported form the Northeast China, including Shandong, Henan, Hebei, Anhui and Jiangsu Area. Figure5d shows the air masses on 29 August received some long-range transport effects from East China and South China. Figure 6 shows that sampling area received the effects from local and long-range transport air masses. Meanwhile, various meteorological factors are also important effective factors. It is necessary to do more work to get more details about air pollution of Wuhan in our further research.  Figure 5a shows that air masses on 16 February may be mainly from northeast and southeast. Figure 5b shows that air masses on 20 February were main from the middle-lower Yangtze River District. The other vital impact factors may be continuous raining. Figure 5c shows that air masses on 18 August were transported form the Northeast China, including Shandong, Henan, Hebei, Anhui and Jiangsu Area. Figure 5d shows the air masses on 29 August received some long-range transport effects from East China and South China. Figure 6 shows that sampling area received the effects from local and long-range transport air masses. Meanwhile, various meteorological factors are also important effective factors. It is necessary to do more work to get more details about air pollution of Wuhan in our further research.  In both winter and summer sampling periods, the CR for adults by Cr and Ni showed that Cr, Ni and As elements in PM2.5 may pose great health risks to the residents in this area. The TCR values of these elements showed that-per million-about 6.94 children and 46.5 adults may risk getting cancer via the inhalation system under this winter atmospheric surroundings, while about 5.41 children and 36.6 adults may be at risk of cancer via the inhalation system under summer atmospheric surroundings. These quantities were greater than those reported in Nanjing, China (Children, 1.32, adults, 5.29) [17] and lower than that in Zhengzhou [12]. Moreover, Wang (2020) studied health risks by heavy metal element in atmosphere of Shanghai and found that they were distributed in finer particles such as PM1.1 [11]. Compared with kinds of crustal elements, these heavy metal elements were only a few tens of nanograms or even less in PM2.5 (μg), but it can cause non-negligible health risks.

Conclusions
In this article, eight water-soluble ionic species and 17 elemental species in PM2.5 collected in central area of Wuhan in February and August of 2018 were measured to analyze the variation of concentrations, possible source and health risk assessment. PM2.5 in winter was about twice that of the summer levels; this result was consistent with the previous studies. Eight water-soluble ionic species were about 1/3 of the PM2.5, with secondary ionic aerosols the dominant content (>85%) during both of the two sampling periods. The ionic balance (<1.00) indicated that there may be some unmeasured ionic contents, such as of CO3 2− , HCO3 − , PO4 3− and other anions. Higher ratios of NO3 − /SO4 2− occurred in winter, indicating that vehicle emissions were an important anthropogenic source. Sulfur oxidation ratio (SOR), NOR and the equivalent ratios of (NO3 − + SO4 2− ) to NH4 + showed the different chemical changes of secondary aerosol behaviors between two different periods. Higher concentration of K element and EFs of K during the winter sampling periods indicated the more human activities such as biomass burning and fireworks in winter. The EFs and elemental ratios showed that coal combustion and vehicle emissions were the main anthropogenic sources of PM2.5 in both winter and summer. Backward trajectories of air masses indicated that air mass would get the long-range transport effects from various regions. Carcinogenic risk evaluated by several heavy

Health Risk Assessment Risk by Heavy Metal Elements in PM 2.5
It has already reported that kinds of trace elements in PM 2.5 pose relative human health risks. It could lead to human dysfunction and various illness because PM 2.5 could be inhaled via nose and mouth everywhere [13,15,46]. According to the risk assessment guidance from USEPA [47] and some reported articles [13,15,46], health risk assessment could be divided into 2 groups: non-carcinogenic risk (by V, Cr, As, Mn, Cd, Ni) and carcinogenic risk (by Cr, As, Pb, Cd, Ni). Health risk assessment posed by heavy metals in PM 2.5 referred from USEPA are shown as the following [13,15,46]: The exposure concentration (EC) is the concentration which is based on the 'reasonable maximum exposure '(C, the upper bound of the 95% confidence internal for the average metal concentration); exposure time (ET, 6 h/day), exposure frequency (EF, 350 days/year); exposure duration(ED, children: 6 years, adults: 24 years); average time (AT, for noncarcinogens, AT = ED × 365 days × 24 h/day; for carcinogens, AT = 70 years × 365 days/year × 24 h/day). The hazard quotient (HQ) is the noncancer risk of a single contaminant by means of exposure. RfCi is the inhalation reference concentration (mg/m 3 ) below which adverse noncancer effects are unlikely to occur (RfCi: V, 0.0001 mg/m 3 ; Cr, 0.0001 mg/m 3 ; As, 0.000015 mg/m 3 ; Mn, 0.00005 mg/m 3 ; Cd, 0.00001 mg/m 3 and Ni, 0.00005 mg/m 3 ). The hazard index (HI) is equal to sum of HQ of each heavy metal to assess the overall potential of noncarcinogenic effects posed by various heavy elements. When the HQ (HI) > 1.00, there may be an adverse health effect which needs to be paid more attention. If the HQ (HI) < 1.00, the noncancer health effect is believed to be not significant and may be neglected at certain times [13,15,46]. ICR is the inhalation unit risk posed by each heavy metal element (ICR, Cr, 0.012 m 3 / µg; As, 0.0043 m 3 / µg; Cd, 0.0018 m 3 / µg; Pb, 0.00008 m 3 / µg and Ni, 0.00024 m 3 / µg) while ICR and RfCi of each heavy metal element was cited from regional screening level in resident air supporting tables (http://www.epa.gov/region9/superfund/prg/). The carcinogenic risk (CR) represents the individual cancer number among a certain number of people posed by a single contaminant (1 × 10 −6 , 1 in million) [48]. According to the USEPA's management, acceptable or tolerable CR risk links are between 1 × 10 −6 and 1 × 10 −4 , while CR over 1 × 10 −4 are considered as unacceptable. A CR below 1 × 10 −6 is considered not to pose any significant health effects [16].
Moreover, the non-carcinogenic and carcinogenic risks from trace element in PM 2.5 at each sampling case are summarized in Table S6. Figure 4 shows average hazard quotient (HQ) values for children and adults of V, Cr, As, Mn, Cd and Ni element, and HI values in winter and summer sampling periods. It was found that HI values for children and adults showed major effects from Mn element in PM 2.5 . The hazard index values were 0.82 in winter and 0.63 in summer. Both were under 1.00, which indicated that there were not serious non-carcinogenic risks in the sampling area during the sampling periods. Higher HQ values of Mn in PM 2.5 was also reported in Zhengzhou, Luoyang and Pingdingshan in Henan area [12] and Baotou [43]  In both winter and summer sampling periods, the CR for adults by Cr and Ni showed that Cr, Ni and As elements in PM 2.5 may pose great health risks to the residents in this area. The TCR values of these elements showed that-per million-about 6.94 children and 46.5 adults may risk getting cancer via the inhalation system under this winter atmospheric surroundings, while about 5.41 children and 36.6 adults may be at risk of cancer via the inhalation system under summer atmospheric surroundings. These quantities were greater than those reported in Nanjing, China (Children, 1.32, adults, 5.29) [17] and lower than that in Zhengzhou [12]. Moreover, Wang (2020) studied health risks by heavy metal element in atmosphere of Shanghai and found that they were distributed in finer particles such as PM 1.1 [11]. Compared with kinds of crustal elements, these heavy metal elements were only a few tens of nanograms or even less in PM 2.5 (µg), but it can cause non-negligible health risks.

Conclusions
In this article, eight water-soluble ionic species and 17 elemental species in PM 2.5 collected in central area of Wuhan in February and August of 2018 were measured to analyze the variation of concentrations, possible source and health risk assessment. PM 2.5 in winter was about twice that of the summer levels; this result was consistent with the previous studies. Eight water-soluble ionic species were about 1/3 of the PM 2.5, with secondary ionic aerosols the dominant content (>85%) during both of the two sampling periods. The ionic balance (<1.00) indicated that there may be some unmeasured ionic contents, such as of CO 3  showed the different chemical changes of secondary aerosol behaviors between two different periods. Higher concentration of K element and EFs of K during the winter sampling periods indicated the more human activities such as biomass burning and fireworks in winter. The EFs and elemental ratios showed that coal combustion and vehicle emissions were the main anthropogenic sources of PM 2.5 in both winter and summer. Backward trajectories of air masses indicated that air mass would get the long-range transport effects from various regions. Carcinogenic risk evaluated by several heavy metal elements in PM 2.5 showed that these elements with light concentration also played a great role to health risk. Heavy metal elemental contents were relatively stable in winter and summer. The CR, As, Ni, Cd, Mn generated from anthropogenic sources are considering as the significant harmful metal elements in PM 2.5 in Wuhan area. Even with a good air quality during the summer sampling period in Wuhan, heavy metal elements also played serious carcinogenic risks to the residents. Otherwise, PAHs, dioxin and complex bioaerosol are the important contents in PM 2.5 [18]. All of them prompted us to pay more attention to air quality, especially for its various components, interaction and possible hazards. Moreover, the toxicity and health effects of other critical contents in atmospheric particle research is one of our main works in the future.
Supplementary Materials: The following are available online at http://www.mdpi.com/2073-4433/11/7/760/s1. Table S1: Environmental information and contents urban area of Wuhan during winter and summer sampling periods. Table S2: Separate elemental content in each sampling case during winter and summer sampling periods. Table S3: Separate ionic content in each sampling case during winter and summer sampling periods. Table S4: Pearson's correlation matrix between ions, PMs, meteorological factors and other pollutants.

Conflicts of Interest:
The authors declare no conflicts of interest.